ESMO 2026: big pharma and China raised the bar; Genmab matched it, most other European biotechs brought early data
At ESMO 2026 in Madrid, 23–27 October, none of the 12 late-breaking trials in the three Presidential Symposia came from a European biotech: eight were sponsored by big pharma (AstraZeneca and Merck three each, Amgen and Roche one each), two by Chinese companies (Akeso and Hansoh) and two by academic groups. Several of them set new standards or validated approaches that companies on the 5mn Research list are pursuing, in small cell lung cancer, in vaccines against melanoma relapse and in PD-(L)1×VEGF bispecifics. Among European biotechs, Genmab matched that level with two Phase 3 survival wins in the same week, announced by press release rather than at the congress, while BioNTech’s Phase 3 in breast cancer showed only a modest benefit and Immatics, BioInvent, OSE and Nanobiotix brought single-arm or small randomised data that will now be judged against a higher bar. The first two tables cover the Research list and its read-across; the third ranks the most important data from other companies.
| Company · drug | Setting | Design | Key result |
|---|---|---|---|
| Genmab · Rina-S targets FRα · ADC · press release 23 Oct, not at ESMO | Platinum-resistant ovarian, any FRα level | Ph3 RAINFOL-02 vs chemo, n=584 | PFS HR 0.29; OS HR 0.43 |
| Genmab · petosemtamab + Keytruda targets EGFR×LGR5 · bispecific · press release 28 Oct, not at ESMO | 1L head and neck, PD-L1-positive | Ph3 LiGeR-HN1 vs Keytruda, first interim | OS HR 0.50 |
| Genmab · petosemtamab + Keytruda targets EGFR×LGR5 · bispecific | 1L head and neck | Ph2, n=43 | ORR 63%; mOS 28.2 mo |
| Genmab · petosemtamab + FOLFOX or FOLFIRI targets EGFR×LGR5 · bispecific | Metastatic colorectal, 1L and 2L | Ph2, n=79 | ORR 75% (1L), 67% (2L)* |
| BioNTech · pumitamig + docetaxel targets PD-L1×VEGF-A · bispecific, with BMS · late-breaker | 2L lung (NSCLC) after immunotherapy and chemo | Ph2, single arm, n=30 evaluable | cORR 36.7%; mPFS 6.9 mo |
| BioNTech · pumitamig + elfetabart drozuntecan targets PD-L1×VEGF-A + B7-H3 · bispecific + ADC, with BMS and DualityBio | Advanced solid tumours | Ph2 | Figures n.a. in release |
| BioNTech · trastuzumab pamirtecan targets HER2 · ADC, with DualityBio · press release 26 Oct | HR+/HER2-low breast | Ph3 DYNASTY-Breast02 vs chemo, n=541 | PFS goal met, modest gain; more ILD |
| Immatics · IMA203CD8 targets PRAME · TCR-T cell therapy | Advanced solid tumours | Ph1, n=87 evaluable | cORR 41%; mDOR 8.4 mo |
| Immatics · IMA402 targets PRAME · TCR bispecific | Advanced solid tumours | Ph1, n=40 | cORR 25%; melanoma 32% |
| BioInvent · BI-1808 + Keytruda targets TNFR2 · antibody | Platinum-resistant ovarian, median 4 prior lines | Ph1/2, single arm, n=34 | ORR 24%; mPFS 4.1 mo |
| OSE · Tedopi + docetaxel or nivolumab targets TUMOUR ANTIGENS (HLA-A2) · vaccine | 2L lung (NSCLC) after chemo-immunotherapy | Randomised Ph2, open-label, n=106 | 1-year OS 47% and 46% vs 29% |
| Nanobiotix · JNJ-1900 (NBTXR3) + radiotherapy + anti-PD-1 targets RADIATION DOSE IN THE TUMOUR · nanoparticle, J&J | Anti-PD-1-resistant melanoma | Ph1, single arm, n=24 | ORR 41.7%; mOS 17.8 mo |
▲ randomised trial met its main goal; ▼ randomised trial met its goal with limited benefit; no mark: single-arm or dose-finding data. ADC = antibody-drug conjugate; ORR = response rate (cORR = confirmed); mPFS, mOS, mDOR = median progression-free survival, overall survival, duration of response; HR = hazard ratio vs the control arm. Genmab: RAINFOL-02 enrolled patients regardless of FRα level; medians, response rates and safety rates n.a., to be shown at a future meeting. LiGeR-HN1: first interim analysis, response odds ratio 3.17; patient numbers and rates n.a. Head and neck Phase 2 (presented at ESMO): 43 evaluable of 45 treated, median follow-up 25.5 months, complete responses 19%, mPFS 9.0 months, no grade 5 events. *Colorectal: confirmed and unconfirmed responses, median follow-up 5.9 (1L) and 7.1 months (2L); Phase 3s LiGeR-CRC1 and LiGeR-CRC2 enrolling. BioNTech: ROSETTA Lung-107 enrolled 59 (cORR in 30, mPFS in 50), no control arm, grade 3 or higher treatment-related side effects 18%. Pumitamig + elfetabart drozuntecan (B7-H3 ADC): the ESMO release gives no figures; in small cell lung cancer at WCLC (Sep 2026) the combination had ORR 70.4% in 71 patients. DYNASTY-Breast02: HR and medians n.a.; lung inflammation (ILD) more frequent than in earlier trials; next steps open. Immatics: IMA203CD8 ORR 51% (46/90) including unconfirmed responses, gynaecological cancers at higher doses cORR 55% (11/20), mPFS 5.8 months; IMA402 PRAME-high cORR 32%, mDOR 12.5 months, no ICANS. BioInvent: 1 complete and 7 partial responses; part C, randomised with Keytruda and paclitaxel, from Q1 2027. OSE: Combi-TED is an academic trial (FoRT) in HLA-A2-positive patients without EGFR, ALK or ROS1 alterations, with three arms (Tedopi + docetaxel, Tedopi + nivolumab, docetaxel alone); serious treatment-related side effects 17%, 24% and 17%; median OS n.a. Nanobiotix: one injection into a single tumour, then stereotactic radiotherapy and anti-PD-1; 4 complete responses; mPFS 6.9 months. Checked, no ESMO 2026 data found: Transgene (TG4050 three-year data by press release, 2 Sep 2026), Immunocore (PRAME data in ovarian and lung cancer guided for H2 2026), PharmaMar, Philogen, Innate, Bicycle and Molecular Partners (next data at NACLC and ASH in December).
| Trial · drug | Setting | Result | Concerns |
|---|---|---|---|
| DeLLphi-305 · Imdelltra + Imfinzi targets DLL3×CD3 · Amgen, AstraZeneca · LBA3 | 1L maintenance small cell lung, n=563 | OS HR 0.42; 21.1 vs 12.7 mo | ▼ PharmaMar: same setting as Zepzelca + Tecentriq (IMforte OS HR 0.73) |
| IDE849 · zoci targets DLL3 · ADCs · IDEAYA with Hengrui; Zai Lab · Ph1 | Small cell lung, 2L and 1L | ORR 72.5% (2L); cORR 81.8% (1L) | ▼ PharmaMar: DLL3 drugs moving into earlier lines |
| INTerpath-001 · intismeran + Keytruda targets THE PATIENT’S OWN NEOANTIGENS · mRNA · Merck, Moderna · LBA1 | Melanoma after surgery, stage IIB–IV | Relapse HR 0.58; 2-year 82.2% vs 70.0% | ▲ Transgene: same approach as TG4050; BioNTech’s autogene cevumeran behind |
| HARMONi-GI1 · ivonescimab + chemo targets PD-1×VEGF · Akeso · LBA8 | 1L biliary, n=682, vs Imfinzi + chemo | OS HR 0.75; 14.9 vs 12.1 mo | ▲ BioNTech: same class as pumitamig |
| ARTEMIS-011 · risvutatug rezetecan targets B7-H3 · ADC · Hansoh; GSK outside Greater China · LBA9 | Osteosarcoma after 2 or more lines, China, n=122 | PFS HR 0.33; 5.5 vs 1.4 mo | ▲ BioNTech: same target as elfetabart drozuntecan; ▼ PharmaMar: GSK Ph3 in late-line small cell lung |
| Sofe-M · ZW191 targets FRα · ADCs · Lilly; Zymeworks · Ph1 | Ovarian: platinum-sensitive (Sofe-M + bevacizumab); platinum-resistant (ZW191) | cORR 79.3%; 55.3–64.5% | Genmab: FRα rivals to Rina-S, still in Phase 1 |
| BELLA · relacorilant + nab-paclitaxel + bevacizumab targets GLUCOCORTICOID RECEPTOR · Corcept · Ph2 | Platinum-resistant ovarian, n=95 | ORR 59%; mPFS 7.2 mo | BioInvent, Genmab: higher bar for chemo-based combinations |
| Panku-Lung01 · OptiTROP-Lung06 targets EGFR×HER3; TROP2 · ADCs · BMS with SystImmune; Merck with Kelun · China Ph3 | Lung (NSCLC): EGFR-mutant after TKI; 1L, PD-L1 below 1% | mPFS 11.1 vs 5.7 mo; PFS HR 0.49 | OSE: ADCs moving into the lines around 2L lung cancer |
| VOLGA · Imfinzi + enfortumab vedotin around surgery targets PD-L1 + NECTIN-4 · ADC · AstraZeneca · LBA2 | Muscle-invasive bladder, cisplatin-ineligible, n=462 | OS HR 0.58; 2-year 78% vs 69% | Innate: more patients will have had enfortumab vedotin, the group IPH4502 enrols |
First column: ▲ positive data for the company that ran the trial. Last column: ▲ supports the approach of the company named; ▼ raises the bar for it; no mark: mixed or context. Cross-trial comparisons are information only: trials differ in patients, control arms and follow-up. LBA = late-breaking abstract in a Presidential Symposium. DeLLphi-305: open-label, after induction with Imfinzi and chemotherapy; PFS HR 0.34 (7.4 vs 1.9 months); grade 3–4 treatment-related side effects 20.9% vs 10.8%; published in NEJM. IMforte (Lancet, June 2025): maintenance after Tecentriq-based induction, 483 randomised, OS 13.2 vs 10.6 months, PFS HR 0.54. DLL3 ADCs: single-arm Phase 1, mostly in China (IDE849 2L n=51, confirmed ORR 60.8%; zoci with Tecentriq 1L n=33). INTerpath-001: double-blind; no difference in the first six months; overall survival immature (HR 0.77). Transgene’s TG4050: no relapse among 16 vaccinated patients against 3 of 16 unvaccinated at a median 41 months (Phase 1 part). HARMONi-GI1: China only, interim; PFS HR 0.61; serious side effects 43.4% vs 43.7%; pumitamig is in seven Phase 3s, none in biliary cancer. ARTEMIS-011: randomised 2:1; overall survival immature (HR 0.51); grade 3+ treatment-related side effects 93% vs 83%; GSK’s programme for the drug includes Phase 3s in late-line small cell lung and prostate cancer; BioNTech’s B7-H3 ADC is developed with DualityBio. Sofe-M: n=29 evaluable; ZW191: high-grade serous cancers, by dose (n=80 in the dose-optimisation part); both with short follow-up and no PFS yet. Panku-Lung01 and OptiTROP-Lung06 are China-only; Combi-TED excluded patients with EGFR mutations. VOLGA: Imfinzi with enfortumab vedotin before surgery and Imfinzi after, against surgery with or without approved adjuvant treatment; event-free survival HR 0.51 (2-year 64% vs 45%); grade 3+ side effects 64% vs 44%; US priority review since September 2026. Innate’s IPH4502 (Nectin-4 ADC, exatecan payload) enrols urothelial cancer relapsed or refractory to enfortumab vedotin, first data at ENA on 18 Nov 2026; Bicycle deprioritised its Nectin-4 drug zelenectide in March 2026. Approved benchmarks in the settings of the first table: platinum-resistant ovarian — Keytruda + paclitaxel (KEYNOTE-B96, US approval Feb 2026: PFS HR 0.72, OS HR 0.76) and Elahere in FRα-high disease (OS HR 0.67); anti-PD-1-resistant melanoma — lifileucel (Amtagvi, US Feb 2024: ORR 31.5%, n=73) and RP1 + Opdivo (Tudriqev, US Aug 2026: ORR 24.2%, n=91), both single-arm accelerated approvals; 1L head and neck — Keytruda alone or with chemotherapy, the control arm in LiGeR-HN1.
Dates from company releases; hollow = expected, as guided by the company. NANORAY-312 interim: Nanobiotix estimate of October 2025. Filing plans for RAINFOL-02 and LiGeR-HN1 n.a.
Beyond the Research list, two kinds of advance stood out. Antibody-drug conjugates (ADCs), which carry a chemotherapy payload to the cancer cells that bear their target, extended survival in gastric and endometrial cancer and delayed progression in osteosarcoma and lung cancer: all five such ADCs in this note that are not yet approved in the US were discovered by Chinese companies (Kelun, KYM, Innovent, Hansoh, SystImmune) and licensed to, or co-developed with, Merck, AstraZeneca, Takeda, GSK and BMS. Newer targeted drugs and immunotherapy also changed what treatment can do: Roche’s divarasib beat the approved KRAS G12C drugs head-to-head on survival, and GSK’s dostarlimab alone cleared detectable cancer for at least a year in 68% of patients with a DNA-repair-deficient rectal cancer, by the investigators’ assessment, without surgery.
| Trial · drug | Setting | Result |
|---|---|---|
| 1 · Krascendo 1 · divarasib targets KRAS G12C · Roche · LBA5 | Lung (NSCLC), previously treated, vs sotorasib or adagrasib, n=338 | OS HR 0.56; 17.7 vs 12.0 mo |
| 2 · CLARITY-Gastric01 · sonesitatug vedotin targets CLDN18.2 · ADC · AstraZeneca, from KYM · LBA7 | Gastric, 2L or later, n=532 | OS HR 0.58; 10.6 vs 6.4 mo |
| 3 · G-HOPE-001 · arcotatug tavatecan targets CLDN18.2 · ADC · Innovent; Takeda outside Greater China · LBA | Gastric, after 2 to 4 lines, China and Japan, n=464 | OS HR 0.57; 10.2 vs 6.9 mo |
| 4 · TroFuse-005 · sacituzumab tirumotecan targets TROP2 · ADC · Merck, from Kelun · LBA11 | Endometrial after platinum and immunotherapy, n=776 | OS HR 0.68; 19.0 vs 12.1 mo |
| 5 · AZUR-1 · dostarlimab alone targets PD-1 · GSK · Ph2, single arm | Locally advanced rectal, dMMR/MSI-H, n=144 evaluable | 68% with no detectable cancer at 12 months, no surgery |
| 6 · SAFFRON · Tagrisso + savolitinib targets EGFR + MET · AstraZeneca, HUTCHMED · LBA6 | Lung, EGFR-mutant and MET-altered, after Tagrisso, n=338 | OS HR 0.68; 17.8 vs 12.9 mo |
| 7 · KEYNOTE-C93 · Keytruda alone targets PD-1 · Merck · LBA12 | 1L endometrial, dMMR, vs chemo, n=299 | PFS HR 0.57; 38.2 vs 8.4 mo |
| 8 · DASL-HiCaP · darolutamide + hormone therapy + radiotherapy targets ANDROGEN RECEPTOR · University of Sydney (ANZUP); Bayer drug · LBA4 | Very high-risk localised prostate, n=1,106 | Metastasis-free HR 0.42; 4-year 95% vs 88% |
| 9 · NATALEE, 6 years · Kisqali + hormone therapy targets CDK4/6 · Novartis | Early HR+/HER2− breast, high risk, n=5,101 | OS HR 0.81; iDFS HR 0.74 |
| 10 · MOIO · immunotherapy every 3 months targets DOSE DE-ESCALATION · UNICANCER · LBA10 | Solid tumours responding after 6 months of immunotherapy, n=340 | Not non-inferior: PFS HR 1.35; OS HR 2.00 |
Ranked by 5mn’s judgement of importance: whether patients lived longer and by how much, whether the result changes current treatment, and how new the drug or approach is. AZUR-1 is ranked above randomised trials because it lets some patients avoid surgery and has already led to a US approval; NATALEE confirms a benefit for a drug already approved in this setting. ▲ randomised trial met its main goal; ▼ randomised trial failed its main goal. Krascendo 1: head-to-head against the approved KRAS G12C drugs; PFS HR 0.58 (9.8 vs 5.7 months); confirmed ORR 58% vs 32%; treatment-related discontinuations 4.8% vs 11.7%; published in NEJM. CLARITY-Gastric01: CLDN18.2 in at least 25% of tumour cells; third line or later OS HR 0.47 (10.0 vs 5.0 months); PFS not significant (HR 0.77); grade 3+ treatment-related side effects 36.4% vs 44.8%; licensed in 2023 from KYM Biosciences, a joint venture of Keymed Biosciences and Lepu Biopharma. G-HOPE-001: CLDN18.2 in at least 75% of cells; PFS HR 0.45 (7.1 vs 3.1 months); grade 3+ treatment-related side effects 47.4% vs 55.9%. TroFuse-005: vs doxorubicin or paclitaxel; PFS HR 0.59; grade 3+ treatment-related side effects 54.4% vs 41.4%; US decision due 15 Jun 2027. AZUR-1: investigator-assessed (98/144); by independent review 47% (54/116), still maturing; no relapse among responders after a median of more than two years; US approval in October 2026. SAFFRON: vs platinum chemotherapy; PFS HR 0.55; US filing made. KEYNOTE-C93: overall survival not significant (HR 0.77); grade 3+ treatment-related side effects 18.1% vs 50.7%. DASL-HiCaP: investigator-led; overall survival not significant at the interim (HR 0.88). NATALEE: absolute gain at six years 5.3 points in invasive disease-free survival (iDFS) and 2.1 points in overall survival; OS HR 95% CI 0.66–0.99. MOIO: immunotherapy every 3 months vs standard intervals after a response at 6 months; the authors argue against uniform de-escalation. Not ranked here because they are in the read-across table: DeLLphi-305, INTerpath-001, HARMONi-GI1, ARTEMIS-011, VOLGA, Panku-Lung01 and OptiTROP-Lung06.
In plain words
What happened. At ESMO, Europe’s main cancer congress, large drug companies and Chinese groups showed several drugs that helped patients live longer or stay free of cancer longer, for example in small cell lung cancer and in skin cancer after surgery. In the same week, Genmab, a Danish company, said that two of its drugs had helped patients live longer in large trials, in ovarian cancer and in head and neck cancer. Many of the new drugs that worked carry a chemotherapy payload straight to cancer cells (antibody-drug conjugates), and most of those were first discovered by Chinese companies.
Why it matters. When a new treatment becomes the standard, a smaller company’s drug has to do better than that new standard, not the old one. Most European biotechs showed early results in small groups of patients, so their next, larger trials will face a higher bar.
The key question Which European programmes still look competitive after the bar moved? Genmab now has randomised survival data in two diseases, but its position in ovarian cancer will be contested: Lilly’s and Zymeworks’ FRα drugs also show responses across FRα levels, though only in Phase 1, and Genmab’s releases give no medians or filing date yet. PharmaMar faces the largest shift: in the maintenance setting where Zepzelca with Tecentriq reached an OS hazard ratio of 0.73 in IMforte, Imdelltra with Imfinzi reached 0.42 in DeLLphi-305, a cross-trial comparison of open-label trials with different induction treatments, while DLL3 drugs are moving into earlier lines and PharmaMar’s LAGOON missed its goal in second line in June. Transgene’s approach gained most: INTerpath-001 is the first randomised Phase 3 win for an individualised neoantigen vaccine, though the benefit appeared only after six months and Transgene’s own randomised data come by end Q1 2028. BioNTech’s pumitamig gains from another randomised survival win for the PD-(L)1×VEGF class but has no randomised data of its own yet, and its HER2 ADC showed a limited benefit. In previously treated melanoma, RP1 with Opdivo was approved in August 2026 on a 24.2% response rate and lifileucel in 2024 on 31.5%: Nanobiotix’s 41.7% in 24 patients needs a randomised test, and Immatics’ Phase 3 SUPRAME, whose control options include lifileucel, reports in H1 2027. In platinum-resistant ovarian cancer, BioInvent’s part C will be measured against Keytruda plus paclitaxel, not Keytruda alone. OSE’s Combi-TED is randomised but small and open-label; ARTEMIA, testing Tedopi alone against docetaxel, reports in early 2028. Beyond the Research list, the five new ADCs with Phase 3 wins in this note were all discovered in China and licensed to large Western groups, which shows where much of the class is now being invented; European ADC developers such as Genmab and Innate compete with that pipeline.
Disclosures: figures marked calc. or est. are 5mn’s own. 5mn does not cover companies it advises, and 5mn and its analysts hold no positions in the companies mentioned. This note is general information, not investment advice or a recommendation to buy, sell or hold any security. Full terms: 5mn.ch/disclaimer.